PART I11. | THE MECHANISMS OF FLOWERS. 453 
upper lip, a triangular surface, which does not bend, but gets 
pushed outwards by the bee’s head. But the points where the 
evolute edges and the thickened bands meet are on each side of 
the-wide part of the opening, and as they separate more and more 
from one another, the superior angle (¢, g, e, 1) of the small part 
of the entrance above this will be very greatly enlarged, for the 
sides bounding it are very short, and are forced apart as much 
as the long sides of the lower opening. The small processes (g, 1, 2) 
which before nearly met, and which held together the two halves 
of the pollen-receptacle, are forced apart; the anthers, which have 
a tension outwards in consequence of the peculiar curvature of 
their filaments (4, 7, 3),1 are freed from the force that held them 
together below, while they remain fastened together above ; they 
therefore flap apart, and let a little pollen fall on the bee’s head, 
exactly on the spot which came in contact with the stigma scarce 
a second before. The pollen is saved from being scattered at the 
sides by vertical hairs on the longer stamens (/, 3),? which cover 
the space between the upper and lower anthers on each’ side, and 
project slightly beyond the lower edges. 
Bombus hortorum, L. 2 (20 to 21 mm.), needs to thrust its head 
a very little way down after inserting it into the wide entrance ; 
but the shorter a bee’s proboscis is the more must it force its 
head down in the upper lip, and if there is not space enough above 
the platform of the lower lip, the latter can be pressed down 
3 to 4 mm., as far as b, 2; so that bees with a proboscis only 
10 mm. long may reach the honey. When the bee flies away, the 
lower lip springs back into its former position, and the whole 
mouth of the flower resumes its original state. 
The flower is in this way adapted for all our native species 
of Bombus and Anthophora, except B. terrestris and small workers 
of a few other species; but the length of the tube excludes all 
smaller bees, which if the tube were shorter might carry off the 
honey without touching the stigma; the hooded upper lip guards 
the pollen from flies and other insects; but the tube is liable 
to be bitten through and robbed of its honey by some humble- 
bees. Such robbery does little or no harm, for bees still visit 
1 T have never noticed the anthers adhering at all to the inner surface of the 
cain as Dr. Ogle describes. Such an adhesion, if it exists, must be very slight 
indeed. 
* According to Dr. Ogle (No. 632, p. 46), the lower anthers are held together by 
the pressure of these hairs upon the wall of the corolla. To act in this way the 
hairs would have to be directed obliquely outwards, They are not so, but lie parallel 
to the median vertical plane of the flower. . 
8 Dr. Ogle is mistaken in thinking (No. 632, p. 180) that the wide calyx is suffi- 
cient to protect Pedicularis silvatica from robbery on the part of short-tongued bees. 
